8
$\begingroup$

I have a z-axis which follows schematically the same principle as the makerbot one's (threaded rod and two leading rods with linear ball bearings carrying a level). It is from an old experimental lab doing physics or chemistry.

The axis move gorgeously about 5 cm, but then it get's stuck on either sides of this way. Both driving threaded rod and the leading rods seem perfectly fine and should be able to allow for further movement. This is as far as I can see by eye.

Where should I look to find further issues and how could I improve the performance? Do I need to take the construction apart?

$\endgroup$
3
$\begingroup$

There could be a few issues at play.

  1. The smooth rods are not parallel which is causing the bearings to bind the further you go up.
  2. Part of the thread is damaged not allowing it to pass through the nut.
  3. The threaded rod is bent significantly to where it either doesn't pass through the nut or bind the assembly. (Is the end of the threaded rod opposite the motor constrained?)
  4. Even less likely is that the motor could be damaged. Since you said it moved 5mm I would have to assume that is several revolutions of the motor so this is unlikely but possible.

The easiest way to find the problem would be to unhook/remove the threaded rod and see if you can move the carriage up and down the rods by hand, if so you just narrowed the problem down significantly.

| improve this answer | |
$\endgroup$
  • $\begingroup$ I had to remove the coupling to the threaded rod to find out that the smooth rod arrangement was not the source of the problem. The threaded rod was probably not used for some years and probably never at all in certain areas. A thorough cleaning beyond visible clanliness brought everything back to life. $\endgroup$ – kamuro Jul 4 '16 at 7:54
5
$\begingroup$

X stage binding like this is almost always caused by parallelism issues with the rods and/or screw. The two-rod-plus-screw arrangement is quite over-constrained and thus requires good alignment to move smoothly.

Some basic troubleshooting steps:

  1. Make sure the screw is not constrained at both ends. It is very difficult to manufacture a screw that is perfectly straight and then mount it so precisely that it can rotate with zero runout. Screws in light-duty linear motion applications should be allowed to "wobble" freely so any bend or runout doesn't apply side-load forces to the Z stage. This can be accomplished by putting a misalignment-tolerant coupler (like an Oldham coupler) on the driven end, and/or leaving the non-driven end free without any support bearing. A motor mount with a small amount of compliance (like a rubber stepper damper) can help when the screw is captive to the motor.
  2. Lubricate the screw and bearings and make sure everything is clean and in good condition.
  3. Allow the Z carriage to self-align the rods and screw. The proper technique for this will depend on the Z stage, but the basic idea is to loosen the rod and screw mounts on either end, run the Z carriage back and forth a few times to push the rods into position, and then only re-tighten the top and bottom hardware when the carriage is at that end. It may also be necessary to loosen the screw nut and bearings on the carriage to get everything aligned properly and smooth-running, but that does not enforce parallelism like loosening the rod and screw mounts, so is really a secondary step. It may be necessary to leave some "float" in one rod or the bearings on one rod (with gravity preloading out any resultant slop) if the hardware has major alignment issues.
  4. If the stage still binds after doing the above, it may be necessary to check if the linear hardware is bent (such as by rolling rods on a flat table) or use a file or Dremel to loosen up the mounts or whatever feature is causing the misalignment.
| improve this answer | |
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.